Measurement of glomerular filtration rate in renal transplant recipients: A comparison of methods

While single injection radionuclide and radio-contrast glomerular filtration rate (GFR) clearance methods are widely used, the relative accuracy of single versus multiple sample techniques continues to be debated. In addition, GFR calculated from the serum creatinine concentration is considered by some to produce results comparable to clearance methods. In this study, 109 patients with stable renal transplant fraction were prospectively evaluated by yearly Cr-51 EDTA clearance as well as by three published formulae used to predict GFR from serum creatinine. Analysis of 362 measurements demonstrated a highly significant correlation between multiple and single point clearance results, as well as the serum creatinine nomograms using least squares regression analysis (P < 0.001). The mean GFR was, however, significantly higher using the Cockcroft and Gault formula; 64 +/- 18 compared with 47 +/- 14 and 50 +/- 14 with other serum creatinine formulae, and 46 +/- 21-50 +/- 17 mL/min per 1.73 m(2) with the three Cr-51 EDTA methods (P < 0.01). However, further statistical analysis using more appropriate methods, including an analysis of difference and least product regression analysis did not support any of the methods tested as reliable alternatives to multiple sample Cr-51 EDTA clearance, because both fixed and proportional bias was noted. In a subgroup of 29 patients evaluated yearly over the 7-year study period, serum creatinine derivations all demonstrated a greater year to year mean fluctuation compared with clearance methods. It is concluded that while each GFR method has similarities, they are not interchangeable. Until clearance methods and serum creatinine formulae are directly compared with inulin clearance with the use of appropriate statistical evaluation, it is recommended that the Chantler 3 sample Cr-51 EDTA method be the method of choice in clinical laboratories.